Ecommerce conversion rate: calculate and improve it

Conversion rate is the metric people talk about most and make the worst decisions with. The reason: as a single figure across all channels, devices and products it is a blend that almost never suggests an action. This article shows how to calculate it correctly, where you have to split it so it means something, and which levers actually move it in a store.

How to calculate conversion rate

Conversion rate = orders ÷ sessions × 100

Example: 40,000 sessions and 800 orders in 30 days gives a conversion rate of 2.0%.

Three things decide whether the number is usable.

Sessions, not users. A customer who visits three times and buys once produces three sessions and one order. Calculate with users instead of sessions and the value comes out higher. Both are legitimate, but you must not mix them, and GA4 offers both.

The same period for numerator and denominator. Combining orders from the store back office with sessions from GA4 across different date ranges or time zones is a common error. In GA4, check the property's time zone setting.

One defined source. The store knows every order, GA4 only the tracked ones, so GA4 figures sit systematically lower. Pick one source as the truth for the trend and stick with it. As an experience value, a gap beyond roughly 10% is worth treating as a tracking problem and fixing before you optimize anything, and the steps are in what to do when GA4 ecommerce tracking is not working. On Shopify specifically, the reasons store revenue does not match GA4 are worth reading before you assume something is broken.

What a good conversion rate is, and why the question rarely helps

Across many stores the values commonly land between 1% and 3%. That range is an experience value, not a target, because it is driven by factors you cannot change in the short term:

• Price level. A store selling $800 bikes will never reach the rate of a store selling $12 coffee. Considered purchases need several visits.
• Share of brand search. People who arrive by typing your store name usually intend to buy. A store with a strong brand share has better numbers for that reason alone.
• Purchase frequency. Consumables with regular customers sit structurally higher than one-off purchases.
• Traffic mix. Cold social traffic converts worse than searches with buying intent. More reach from a weakly converting channel lowers the blended figure even though nothing is broken.

The comparison that is actually useful is always internal: your value against last month, against the same month last year, and most importantly segment against segment.

Where you have to split it

Take the last 90 days and split conversion rate along these four dimensions. In practice you find an outlier in at least one of them that explains a large part of the potential.

An example of the effect: if 60% of your sessions are mobile and the rate there is 1.2% instead of 2.4%, that single number holds more revenue than five new campaigns.

Read the funnel, not the end figure

Conversion rate only tells you that something is missing, not where. For that you need the intermediate steps GA4 delivers through the standard ecommerce events:

1. Product view (view_item) — do visitors reach products at all?
2. Add to cart (add_to_cart) — does the product page convince?
3. Checkout started (begin_checkout) — is the intent there?
4. Purchase (purchase) — does the checkout deliver what the product page promised?

Example: 20,000 product views, 1,800 carts, 900 checkouts started, 630 purchases. Only half of the carts make it into the checkout, so that is the issue, not the product page.

In GA4 the purchase journey and checkout journey reports under Monetization cover this without any setup, and a funnel exploration lets you break the same steps down by device. Shopify shows a simplified version in its own behavior reports: all sessions, sessions with cart additions, sessions that reached checkout and sessions that completed it. For a first look that is often enough. What to do about a weak final step is in the guide to reducing cart abandonment.

Ten levers that actually move conversion rate

Sorted by effort-to-effect as it usually plays out in small and mid-sized stores.

1. Show shipping cost and delivery time early

The most common reason for abandonment is a surprise in the last step. State shipping cost and a realistic delivery time on the product page, not first in the checkout. If you have a free-shipping threshold, show the remaining amount in the cart.

2. Allow guest checkout

A forced account before the first purchase costs orders. Offer the account after the order, not before.

3. Add the payment methods your audience expects

If the payment method your customers are used to is missing, they leave without telling you. Which ones matter for you is visible in the distribution of your existing orders and in the drop-off after add_payment_info. Express payment methods have the extra benefit of handling address and payment in one step.

4. Load time on mobile

Slow pages lose visitors before they see anything. Measure with PageSpeed Insights and work on Core Web Vitals, above all LCP on product and category pages. The store-specific version is in Core Web Vitals for online stores; the official definitions of the metrics are on web.dev.

5. Product images and zoom

Several images, detail shots, a size reference, and for clothing photos on a person. Images replace touching the product, and too few images is one of the most common causes of a weak add-to-cart rate.

6. Availability and stock status

In stock, ships today works better than any discount. The reverse is also true: unclear availability creates abandonment. If something is not available, name a date.

7. Reviews on the product page

Real reviews with text, not just stars. Mixed reviews help too, because they are credible. A page with nothing but five-star ratings convinces less than one with a critical comment in between.

8. Internal search and filters

In a store with more than a few dozen products, internal search is a main route to purchase. Look at what people search for and what returns zero results. That is a list of concrete tasks.

9. Treat category pages as selling pages

Many visitors land on categories, not products. A category page needs an understandable sort order, useful filters and a short explanatory text so people can tell the options apart.

10. Bring in the right visitors

The underrated lever: part of a weak conversion rate is not a store problem but a traffic problem. Queries without buying intent and broadly targeted campaigns deliver sessions that never wanted to buy. Check the search terms report in Google Ads and exclude what does not fit. If that pattern dominates, work through clicks but no conversions in Google Ads.

Checklist: conversion diagnosis in 45 minutes

1. Compare GA4 orders with the store back office for 30 days. Off by much more than 10%? Fix tracking first.
2. Split conversion rate by device over 90 days. Note the mobile-to-desktop gap.
3. Split by channel. Look at brand search separately.
4. Look at the funnel: product view, cart, checkout start, purchase. Mark the biggest drop.
5. List your top ten landing pages by sessions with conversion rate next to each.
6. Walk your own checkout on a phone as a new customer. Note every irritation.
7. Check the Google Ads search terms report for queries without buying intent.
8. Pick exactly one action from all of that and implement it.
9. Measure after at least two weeks or a few hundred orders.

Point 8 is the important one. If you change three things at once, you will not know afterwards which one worked.

Worked example: what an improvement is worth

Starting point: 40,000 sessions, 2.0% conversion rate, $90 average order value, so 800 orders and $72,000 revenue per month.

The rate rises to 2.3%: 920 orders, $82,800. That is $10,800 more per month without a single extra visitor.

For comparison: to get the same effect through traffic you would need 6,000 additional sessions. At a $0.60 cost per click that is $3,600 in ad spend, every month again.

That is exactly why conversion rate often sits ahead of new channels on the priority list. The second lever with a similar effect is average order value, because it works the same way: more revenue from the traffic you already have. How both fit next to your other figures is in the overview of the ecommerce metrics that matter.

When not to work on conversion rate

There are situations where a falling value is not a problem. If you are building a new channel, broadening targeting or buying reach at the top of the funnel, the blended figure has to fall. As long as revenue per session across all channels stays stable and contribution margin grows with it, that is the right development. In ecomando both numbers sit side by side so the connection does not get lost.

And with very small data volumes, conversion rate is simply not meaningful. At 30 orders a month, three orders more or less move it by a tenth of its value. In that case work on obvious errors instead of details, and on reach.

A short word on testing

A/B tests are the correct way to verify a change, and they are also the most commonly misused one. Two practical constraints:

• Volume before duration. You need enough orders per variant, not enough days. Many small stores simply do not generate the volume for a meaningful test in a reasonable time.
• Test one thing. A redesigned product page with five changes tells you that something worked, not what.

If you do not have the volume, the honest alternative is sequential change: one adjustment, at least two to four weeks, note the date, compare the same segment. That is weaker evidence than a test, but it is far better than changing five things and guessing.